TLDR: Global analytics leader FICO has launched its Focused Foundation Model (FFM) for financial services, featuring specialized Generative AI models (FLM and FSM) designed to deliver accurate, auditable, and responsible AI outcomes, significantly improving risk management, compliance, and fraud detection.
Mumbai, India – October 8, 2025 – FICO, a global analytics company, today announced the launch of its groundbreaking Generative AI model specifically tailored for the financial services sector. Named the FICO Focused Foundation Model (FICO FFM), this innovative suite aims to provide accurate, auditable, and responsible AI outcomes, marking a significant advancement in how financial institutions manage risk, ensure compliance, and detect fraud.
The FICO FFM comprises two core offerings: the FICO Focused Language Model (FICO FLM) and the FICO Focused Sequence Model (FICO FSM). These models are designed with a precise, auditable focus on the specific data used to build and task-train them for particular business problems within financial services. This approach stands in contrast to general-purpose large language models (LLMs), which often struggle with the stringent accuracy and compliance requirements of highly regulated industries.
Scott Zoldi, Chief Analytics Officer at FICO, emphasized the strategic shift, stating, “The focused foundation model represents a practitioner’s approach to GenAI in financial services, moving beyond trying to refine universal knowledge models.” He highlighted that these specialized models have demonstrated impressive results, including “38% lifts in compliance adherence use cases” during testing, and more than “35% lifts in world-class transaction analytic models” for areas like fraud detection.
One of the key advantages of FICO’s domain-specific models is their efficiency. Unlike larger general-purpose LLMs, the FICO FFM requires up to 1,000 times fewer computational resources, making them more cost-effective to develop, maintain, and adapt. This efficiency also contributes to a lower barrier to entry for financial institutions looking to leverage generative AI.
Megha Kumar, research vice president of analytics and AI analyst at IDC, noted the transformative potential of these models, stating that Focused Language Models (FLMs), which are essentially similar to small language models (SLMs), are “transforming how GenAI is used in financial risk management and compliance by providing highly accurate, domain-specific insights and reducing misinformation.”
The FICO FFM is architected for trusted and responsible use of GenAI. It incorporates proprietary Trust Scores that provide risk rankings for AI-generated outputs, enabling financial institutions to set specific risk thresholds and mitigate inaccuracies. The FICO FLM is trained using business-specific and task-driven datasets, which significantly lowers the occurrence of ‘hallucinations’ compared to more generalized systems. Meanwhile, the FICO FSM is specifically designed for transaction sequence analysis, capable of identifying complex patterns in payment histories that traditional analytics might overlook, thereby enhancing fraud detection and risk assessment capabilities.
Also Read:
- Nvidia’s Vision: Mumbai to Evolve into a ‘Financial Capital with Intelligence’ by 2028, Driven by AI
- NTT Docomo Business Collaborates with Pier’s to Advance AI Systems for Financial Sector
This launch underscores FICO’s commitment to delivering sustained value through GenAI models across financial services, ensuring transparency, auditability, and adaptability in an increasingly AI-driven landscape.
